If you do not believe software is eating the world, stop reading now, there are better things you can do with your time.
Software is moving from being a supporting component of business to a core component of business. Many companies are using software to build a platform on which their entire business is run. For ex: Uber, AirBnB, Netflix, Amazon.
These early innovators build something crude that’s customized for their use ( For ex: Google’s Borg ). They may open source components of what they built or in some cases, the engineers that built it will start a new startup. The idea here is to take this crude thing that was built and make it general purpose for the market.
This generally worked well so far. This works really well especially in cases where your apps are supporting your business.
If apps are core of your business, this won’t work too well. Why? Because by the time an innovative component is taken by startups and made general purpose, its outdated. If you want to compete, you have to lead the development of new tech, not adopt something that was developed 5 years ago. You also have to keep reinventing for competitive edge. This doesn’t mean developing everything from scratch. You can leverage open source components to build a platform that is custom for your needs. Perhaps, you may open source what you build for lesser companies out there to use hoping to become like you one day.
It is entirely possible that you don’t have the right culture in place to lead innovation as Netflix does. In such a case, you may think adopting a general purpose platform will be a better solution. Unfortunately that isn’t the case. Not addressing culture issues that get in the way of innovation will kill your company. Using a vendor platform to work around culture issues is kicking the can down the road. It will take a little longer to build your own platform, but you are investing in the core of your business.
So, are vendor platforms useless? Not quite. Not every business is going to be software driven business within next decade. The change is happening in phases. If you are in a market where apps play a supporting role, you are better off with a vendor platform since you can get to production faster.
There may come a day when we reach same level of maturity in platforms as we have done with infrastructure. Then you would be better off not building your platform. This is like today where it is an incredibly stupid idea to build your own infrastructure.